f 



SCIENCE AND PRACTICE. 383 



galvanometer equal in every respect, by which, therefore, 

 m = m't and 



To obtain a balance with this arrangement, therefore, it is 

 necessary that the ratio of the capacities of the two jars 







> should be equal to that of the sum of the resistance of the 

 (j 



galvanometer coil, and its shunt divided by the resistance of 

 the shunt ; and for attaining this it suffices to alter the value 

 of r. 



A detailed description of Mr. Yarley's apparatus for 

 this method may be found in Mr. Culley's Handbook. 



89. Final Tests of a Complete Cable. To ascertain whether 

 the electrical conditions of a finished cable correspond with 

 the known conditions of the separate lengths of which it is 

 composed, the conditions of the whole cable must be calcu- 

 lated from the results of the separate tests, and compared 

 with the final test of the complete cable. 



Let the lengths of the coils of the core, tested separately, 



be 1^ l# l# l n , and their sum L ; the insulation 



resistance per knot of the same, after the same time and at 

 the same temperature, r lt r 2) r s , r n , and the resist- 

 ance, per knot, of the whole cable x, then we have the 

 equation, 



whence 



x = 



With a cable carefully made and handled, the measured will 

 agree within a trifle with the calculated value of x. Beyond 

 the errors of observation, the leakage of the joints, which is 

 always great in comparison with that of similar lengths of 

 perfect core, should be the only source of difference. 



The mean value of the copper resistance, per knot, measured 



